Skip to content

A Python based Geo Spatial library to find and merge all geometries that intersects with their neighbor.

License

Notifications You must be signed in to change notification settings

fuzailpalnak/ShapeMerge

Repository files navigation

Shape-Merge

GitHub Python Contributions welcome Downloads

A Python based GIS library for finding and merging all Geometries that intersects with their neighbour. The library will iterate over all the geometries provided in the form of following Inputs, and will look for the neighbours which forms an intersection relationship with its Parent the geometry that looks for intersection is reffered as Parent either through a direct intersection or an intersection link generated via multiple Child geometries which have a relationship associated with its Parent either as a direct neighbour or via mutiple neighbour is reffered to as a Child intersection.

Buy Me a Coffee at ko-fi.com

OutputAnimation

Installation

pip install shape-merge

Requirements

The library uses Rtree which has a dependency on libspatialindex, It is recommend to resolving the dependency through libspatialindex conda

LibSpatialIndex For Linux:

$ sudo apt-get update -y
$ sudo apt-get install -y libspatialindex-dev

LibSpatialIndex For Windows:

Experience is pretty grim for Windows Installation, i used conda for trouble free installation.

Rtree

conda install -c conda-forge rtree

Fiona

conda install -c conda-forge fiona

Inputs

ShapeFile

from shape_merge.merge import ShapeMerge
shape_merge = ShapeMerge()
shape_merge.populate_index_by_fiona(r"path_to_shape_file.shp")

GeoJSON

from shape_merge.merge import ShapeMerge
shape_merge = ShapeMerge()
shape_merge.populate_index_by_geojson(r"path_to_geo_json.geojson")

Iteratively populate the index

from shape_merge.merge import ShapeMerge
shape_merge = ShapeMerge()
for feature in feature_collection:
    shape_merge.populate_index_by_feature(feature)

Feature must be of the following structure:

{'type': 'Feature', 'id': str, 'properties': dict, 'geometry': {'type': 'GeometryType', 'coordinates': list}}

How to run

After Populating the Index, merging is matter of a function call away, execute the following to begin merging:

shape_merge.merge_geometries()

Output

The Output will be a collection, which will contain the merged geometries and the all the ids that were merged together

merged_geoemrty = OrderedDict([(0, {'ids': [ ], 'geometry': {'type': 'GeometryType', 'coordinates': []}})])

Parameters

bounds_buffer :

During rtree index creation the bounds of individual geometry are added with buffer of 0, This param controls on how big the original bounds should grow.

geometry.bounds.buffer(self.__bounds_buffer)

The bounds of the geometry are responsible for finding potential intersecting neighbour i.e everything that lies in the bound is considered as a potential neighbour. A large value of bound value will increase the computational overhead.

geometry_buffer:

Add buffer to geometries while checking if they intersect with each other

geometry_1.buffer(self.__geometry_buffer).intersects(geometry_2.buffer(self.__geometry_buffer))